0

我正在尝试拟合具有 event_type 和 notes（自由文本）列的数据集。在调用 MultinomialNB 模型之前，我处理了文本并将其转换为数组以对其进行矢量化并在提供的代码下方计算 tfidf：

将事件类型从字符串转换为整数以便于处理

ACLED['category_id'] = ACLED['event_type'].factorize()[0]
category_id_ACLED = ACLED[['event_type', 'category_id']].drop_d

uplicates().sort_values('category_id')
category_to_id = dict(category_id_ACLED.values)
id_to_category = dict(category_id_ACLED[['category_id', 'event_type']].values)

文本表示

我还将 notes 和 category_id 转换为特征和标签，如下所示：

tfidf = TfidfVectorizer(sublinear_tf=True, min_df=5, norm='l2', encoding='latin-1', ngram_range=(1, 2), stop_words='english')
features = tfidf.fit_transform(ACLED.notes).toarray()
labels = ACLED.category_id
print(features.shape)

然后我使用特征和标签将数据集拆分为训练和测试集：

X_train, X_test, y_train, y_test = train_test_split(features ,labels, random_state=0)
print('Original dataset shape {}'.format(Counter(y_train)))

输出

Original dataset shape Counter({1: 1280, 2: 819, 0: 676, 3: 593, 4: 138, 5: 53, 7: 50, 6: 21, 8: 10})

由于类不平衡，我使用 SMOTE 来解决少数问题并创建合成副本

应用随机过采样来克服不平衡类

sm = SMOTE(random_state=42)
X_resampled, y_resampled = sm.fit_sample(X_train, y_train)
print('Resampled dataset shape {}'.format(Counter(y_resampled)))

过采样后的输出

Resampled dataset shape Counter({3: 1280, 1: 1280, 2: 1280, 0: 1280, 7: 1280, 6: 1280, 4: 1280, 5: 1280, 8: 1280})

到目前为止一切正常，直到我尝试使用 CountVectorizer() 计算术语频率，如下所示：

count_vect = CountVectorizer()
X_train_counts = count_vect.fit_transform(X_resampled)
tfidf_transformer = TfidfTransformer()
X_train_tfidf = tfidf_transformer.fit_transform(X_train_counts)

输出错误：

'numpy.ndarray' object has no attribute 'lower'

我尝试使用 ravel() 函数来展平数组，但错误仍然存​​在，任何想法，在此先感谢

4

1 回答 1

0

我找到了解决这个问题的方法，而不是使用特征和标签，我直接在数据集上执行了一个子集：

X_train, X_test, y_train, y_test = train_test_split(ACLED['notes'] ,ACLED['event_type'], random_state=0)

然后我在 counVectorizer 之后移动了 SMOTE，因为 SMOTE 有它自己的管道：

将训练集的notes列虚化

count_vect = CountVectorizer()
X_train_counts = count_vect.fit_transform(X_train)
tfidf_transformer = TfidfTransformer()
X_train_tfidf = tfidf_transformer.fit_transform(X_train_counts)

应用随机过采样来克服不平衡类

sm = SMOTE(random_state=42)
X_resampled, y_resampled = sm.fit_sample(X_train_tfidf, y_train)
print('Resampled dataset shape {}'.format(Counter(y_resampled)))

输出

Original dataset shape Counter({'Riots/Protests': 1280, 'Battle-No change of territory': 819, 'Remote violence': 676, 'Violence against civilians': 593, 'Strategic development': 138, 'Battle-Government regains territory': 53, 'Battle-Non-state actor overtakes territory': 50, 'Non-violent transfer of territory': 21, 'Headquarters or base established': 10})
Resampled dataset shape Counter({'Violence against civilians': 1280, 'Riots/Protests': 1280, 'Battle-No change of territory': 1280, 'Remote violence': 1280, 'Battle-Non-state actor overtakes territory': 1280, 'Non-violent transfer of territory': 1280, 'Strategic development': 1280, 'Battle-Government regains territory': 1280, 'Headquarters or base established': 1280})
于 2018-07-03T18:11:27.410 回答